What is Dark Social?

Social Media Article
20 mins

What is dark social, how can we measure it, and what does it mean for your marketing strategy? Join us as we explore one of the trickiest – and most important – topics facing digital marketers today.

The silent majority of social activity

Let’s start with the most fundamental question of all: what is dark social?

“Dark social” refers to all the social media interactions that the public, search engines and marketers like you and I can’t see – such as communications on private social media messaging, email, WhatsApp and SMS.

The term was coined in 2012 by The Atlantic’s Alexis C. Madrigal, in his article, “Dark social: we have the whole history of the web wrong”.

Here’s an abridged version of Madrigal’s concise definition (but honestly read the whole article. It still holds its own a whole six years later!) :

  1. The sharing you see on sites like Facebook and Twitter is the tip of the ‘social’ iceberg. We are impressed by its scale because it’s easy to measure.
  2. But most sharing is done via dark social means like email and IM that are difficult to measure.

All of the following count as “dark social” interactions:

  • Private social media communications on Facebook, Twitter, Instagram, etc.
  • Forwarded emails
  • Messaging apps like Facebook Messenger, WhatsApp, Slack and Snapchat
  • SMS


Why dark social is a problem for digital marketers

One of the most interesting aspects of dark social from a digital marketing perspective is how it plays into web analytics.

Understanding where inbound traffic comes from will be key to any digital marketer’s ability to assess the results of their activities, and thereby to form a future strategy.

Ideally, we can view our web analytics and say things like: “Look – our traffic from Facebook rose during our sponsored post campaign”; or “Reducing our PPC budget seems to have caused a ~10% drop in traffic from paid search over the last week”. Insights like these are simple enough to extrapolate because traffic from traceable sources like paid search and public Facebook campaigns can be measured reliably as a single channel via an analytics platform.

Conversely, traffic from dark social cannot be tracked as a single channel on analytics – and that extends to dark social traffic that really does come from a single channel, such as WhatsApp or Snapchat. Instead, it flows into the “Direct” channel, which also includes visits from users who typed the URL into their address bar, and users who reached the site via a bookmark, as well as various other sources.

This represents a serious obstacle to marketers’ ability to understand how traffic is reaching their web content – especially when considered alongside the fact that dark social sharing is taking place on a truly epic scale.

According to RadiumOne’s report, The Dark Side of Social Sharing, 84% of consumer sharing of marketers’ and publishers’ web content now takes place via dark social channels. By contrast, just 9% is done publicly on Facebook, and 7% publicly on other social media.

In other words, when you look at your social traffic in Google Analytics, you’re only seeing part of the picture. Probably a small part.

Let there be light: proven methods for measuring dark social

We’ll kick off this optimistically titled section with an important dose of realism: there’s no 100% reliable method for measuring dark social traffic, and we have no reason to believe someone’s going to come up with one anytime soon.

That said, there are several established methods you can use to better understand roughly how much of your traffic comes from dark social, and which dark social sources that traffic is coming from. Try these techniques to shine a little light on your web traffic data:

First, look for long URLs

The simplest way to gain an insight into your dark social traffic is to go to the Direct traffic channel in your website analytics and split down the results by landing page.

You’ll probably notice that some of the URLs listed (or perhaps quite a lot of them) are long URLs that users would be unlikely to type out into their address bar.

There’s a high probability that much of the “direct” traffic to these pages actually comes from dark social.

Next, create a dark social segment

In order to draw meaningful conclusions from suspected dark social traffic within the Direct channel, we must weed out as much traffic from other sources as we can.

As Brandwatch’s Joshua Boyd explains in his article on tracking dark social, we can achieve this by carefully configuring a Google Analytics segment. Here’s how:

  1. Set up a direct traffic segment in Google Analytics.
  2. Go to “Advanced” → “Conditions”. Exclude URLs that would likely be typed (i.e. your homepage, and other high-level pages which you consider likelier to be entered directly into a user’s address bar than to be shared via dark social.
  3. Also filter by “Returning Users”, as users are generally likelier to return to a page via truly direct means (i.e. retyping in their address bar or navigating via bookmarks or history), than by using a link shared on dark social repeatedly.

This won’t give you an entirely accurate read on how much traffic you’re getting from dark social, but it’s one of the best solutions we’ve come across.

The curious case of HTTP → HTTPS

Using analytics to identify dark social traffic is a useful exercise – but ultimately a fallible one.

What happens when a web user clicks through from a “secure” site that uses HTTPS to a “non-secure” site using HTTP perfectly illustrates this point.

Analytics tools can’t trace the source when a user navigates from a secure site to a non-secure site, and as such, traffic arriving at your site in this way shows up as direct traffic. In many cases, the destination will be a long URL, and you will, therefore, be prone to inaccurately treating HTTP → HTTPS navigations as dark social traffic.

We haven’t included this example to put your off trying to measure dark social traffic, but rather, to encourage you to treat your findings with a healthy scepticism.

On a side note: could this quirk of HTTPS → HTTP navigations be a motivating factor behind Google’s decision to reward sites that use HTTPS with higher search rankings? As the world’s leading web analytics provider, Google has a vested interest in online activity being traceable.

Use UTM codes to see how dark social traffic plays into specific contexts and campaigns

A more advanced approach to measuring dark social traffic can be achieved by using URLs with UTM parameters in your social media campaigns and social share buttons. Once again, this won’t tell you exactly how much of your traffic is coming from dark social – but it will help you better understand how much dark social traffic is generated by individual links, whether they be share buttons on your website or shortened links in social posts.

A UTM tag is a short text code that can be added to the URL slug of a webpage. They typically look like this:

?&utm_medium=social&utm_source=twitter&utm_campaign=joys-of-wordpress

Or in the context of a full web address, like this:

“https://www.targetinternet.com/the-digital-marketing-podcast-episode-189-the-joys-of-wordpress/?utm_medium=social&utm_source=twitter&utm_campaign=joys-of-wordpress”

In the above UTM example we have three separate UTM tag values for Medium, source and campaign. UTM tags enable marketers to tag useful tracking information onto the links they use to deliver content. The tags most commonly used to track dark social are those shown above: utm_medium & utm_source, but adding a bit of extra information on what the content posted was about via the campaign tag gives you an even deeper level of insight. Let’s briefly talk through how these three tags might be used in a campaign.

So, if you were distributing a link via a Twitter posting campaign, you would set that link’s utm_medium to “utm_medium=social”, and its utm_source to “utm_source=twitter” and utm_campaign to utm_campaign=joys-of-wordpress-podcast

When visitors reach your webpage via that link, it can then be shown in your analytics that they likely arrived directly via Twitter; or as a result of a public or dark social share of the link, stemming from the original tweet. You can thereby gain an enhanced understanding of the campaign’s total impact on traffic to your site, including both public and dark sources. So great. You get a bit more visibility now but be warned. There are two caveats here. Firstly- these trackable links are long and ugly so you may want to spruce them up a little using a nice URL shortening service so they don’t look quite so ugly. ( see below)  Secondly- it’s a lot of work to create all these uniquely trackable links individually for each social media property you post out on every time you post. For a large organisation with a big social following its definitely worth the effort but for smaller organisations with smaller followings, it is a lot of work for not much information back.

If you are after an easy way to build these links then check out Google’s online Campaign URL builder.  Also, be sure to be consistent in how you use each of the fields and keep everything lower case as these codes are case sensitive. Set a medium as social on one link and Social in another and you will forever see the capitalised version separated out in your reports!

Link Shortening Insights

Link shortening services can also help to share some light on dark social sharing . Services like Bit.ly offer both link shortening and tracking, which enables the user to easily create short URLs for their social media marketing campaigns. if you want to add the UTM tracking codes to yoour URL before you shorten it you will get een more detail back ultimately as both Bitly and Google analytics will have additional data to present you but it isn’t essential to use both if you dont feel the need to. Bilty links can  track inbound traffic from each bit.ly link, taking into account both public and dark social sources. To see the extent any bit.ly link has been shared over the last 30 days, simply paste its address into a browser address bar and place a + sign immediately after it. You won’t get to see exactly where the dark sharing came from, but if you are only using that bit.ly link to share to your social media properties then you will have a much clearer indication of how much secret sharing is going on from those channels. Here is an example that we have linked to for you http://bit.ly/2OFYrWm+

If you see the screenshot below you can see the last 30 days. Of the 38 shares this bitly link achieved,  24 of them in this period ( 64%) were shared through non web mediums that will typically show up in analytics as direct to site. That’s actually quite low for us- we see a lot of our bit links getting 70-80% dark shares. Unlike most analytics packages Bit.ly labels that click traffic as ’email SMS or direct’. A bit more helpful and descriptive but essentially they are just not sure entirely where it came from. The cool thing is if we only posted this bit.ly link on social media channels then we know where the original link came from. Of course that all still only works if those practising these dark sharing practices copy and share the shortened bit.ly link. If they copy the full final URL the shoretned links points to, unless there is an additional UTM tracking code in the URL, you really will still be left in the untrackable analytical darkness.

Evidence of Dark sharing example from Bitly analytics

Creating traceable share button links with ShareThis

Share buttons can also be equipped with unique URLs to provide a dark-social-inclusive view of the traffic they send to other web content.

A useful tool for facilitating this approach is something like  ShareThis, which can be used to equip webpages with various types of share button – including social, email and SMS – all of which use unique URLs that will help you reliably trace traffic back to each button. However, despite an army of web usability experts and social gurus pushing site owners for years to employ share buttons to get the shares, in reality very few people use them. They simply copy and paste the URL so their activity isn’t tracked.  By and large, we all know it just isn’t cool to publically share- and that’s really the whole source of the problem!

Turn your findings into an understanding of how dark social plays into your average campaign

Over time, measuring dark social traffic using the methods outlined above will provide an opportunity to calculate the average ratio of visible social sharing to dark social sharing generated by your campaigns.

Simply look at the volume of social traffic generated by your links vs the amount of direct traffic over multiple links and campaigns. This should give a rough idea of how much dark social traffic you can expect to receive, relative to the amount of visible social traffic.

This is well worth doing, as it provides a better basis for estimating the ROI of a planned social campaign than looking solely at the amount of traffic you’d expect from direct social.

The value of dark social traffic

If you have some understanding of dark social, you’re in the minority.

In February 2018, IBM Watson teamed up with Econsultancy to survey over 1,000 European marketers on how dark social will affect their marketing activities (Marketing in the dark, dark social).

Strikingly, 96% of the organisations surveyed said they were unaware of dark social as a risk factor, while the 4% that were aware saw it as a major challenge.

In our view, it’s well worth giving serious consideration to dark social – not only because it likely accounts for a huge share of your social traffic, but also because it likely accounts for the most valuable social traffic.

Whether in terms of product purchases or link click-throughs, a personal recommendation from a trusted peer will almost always create a higher chance of a conversion than a public post (whether by a consumer or a brand).

It therefore pays to encourage dark social sharing of our content – even if we can’t accurately measure it.

How to attract dark social traffic

You may remember Alexis C. Madrigal, the journalist who coined the term “dark social”, from the introduction to this article. His view on how to stimulate dark social sharing is straight and to the point:

“The only real way to optimize for [dark] social spread is in the nature of the content itself. There’s no way to game email or people’s instant messages. There’s no power users you can contact. There’s no algorithms to understand.”

Dark social is where the truly great content gets sorted from the good, and we would agree with Madrigal that the only real way to get content shared is to make it worth sharing.

To that end, we would urge you to dig into your web analytics, using the techniques outlined in this article, and identify the content which attracts a high volume of dark social traffic, compared with the traffic it brings in from other sources. List these high-performing pages, and look for any attributes those webpages tend to share, e.g.:

  • Subject matter
  • Length
  • Content creator
  • Structure

By looking at these and other factors, you may be able to identify some of the key ingredients that encourage your audience to share. And whatever you do, don’t decide based on flawed /misunderstood analytical data that social caren’tls just arent worth it. Dig a bit deeper before you shoot yourself in the foot. Feed these findings into your future content strategy, with a view to increasing the volume of valuable dark social traffic reaching your site.

We may not be able to fully measure dark social, but one thing that’s clear as day is that it matters more than most marketers seem to think it does. Our advice is to understand your dark social traffic as well as you can, and view it not as a threat, but as the greatest opportunity on social today.

Build a ü free personalised ¥ learning plan to see our course recommendations î for you

Free for 30 days

Build a å free personalised ¥ learning plan to see our course recommendations î for you

Free for 30 days